364 THE TSCHERTSCHEN DESERT.
m., that is to say, considerably lower than the sand in the extreme south of the Desert of Tschertschen.
In the Desert of Takla-makan proper, which I crossed in the spring of 1895 from Tschoka-tagh to the Chotan-darja, the arrangement of the sand is excessively irregular. On the extreme west of that part of the desert the dunes plunge steeply down towards the west, south, and south-west, and are drawn out towards the north-north-east and south-south-west, the north and south, and the north-west and southeast. After that I crossed a belt of the desert in which the steep sides were turned to the east and south-east, though this may have been due to a chance local cyclone. Here too there are small level spaces paved with hard clay, but these extend east and west. They are however exceedingly few; we only passed a very small number. They correspond to the bajirs, although they lie in an unusual direction, and it appeared to me, that their floors are as a rule convex rather than concave. Presumably they do not fully equate with the bajirs, but are rather elevations or upswellings of the substratum, around which the sand is therefore always low, for in other parts it reaches altitudes of 5o to 6o m. The accumulations of sand occur again here, though they extend east and west, and are very flat and indistinct. Owing to the structure of the dunes, the journey through the Takla-makan was murderous as compared with that through the Desert of Tschertschen, for in the former they are not arranged in any sort of order whereby we were able to facilitate our march. No matter in which direction we had travelled, the sand would have been equally difficult and equally inconvenient.
Fig. 316. CONVEX BAJIR IN THE TAKLA-MAKAN.
Hence, as the result of my four crossings of this great Central Asian desert, I may say, that its configuration and relief grow more irregular from east to west. In the extreme west, at Ordan-Padschah, where the dunes lie as a rule north and south, these would appear, judging from marks which were shown to me by the sheikhs, to travel annually 4 m. towards the south-east; there the winds blow from April to . June, and the . prevailing wind comes from the north-west. The sharply accentuated relief features which we encountered in the Desert of Tschertschen have already disappeared when one reaches the Kerija-darja. This is a consequence of the wind, which in the whole of East Turkestan blows nowhere with such power and such uniformity as it does in the country of Lop. The modifying effects of the east-north-east wind are here relatively so great that they easily annul the effects of all other winds put together. But in the west, winds from several quarters cooperate together to mould the relief features of the desert into capricious and changing outlines. It is in the wide open Lop country, where the mountains, the Kuruk-tagh on the north and the Astin-tagh on the south, are the nearest to each other that the wind develops its greatest and most concentrated energy. It roars like a tide pouring
IIIF Color HighRes
